The researchers experimentally characterised a screwless, two-plunger micro-injection moulding machine and compared short-shot POM parts with three-dimensional filling simulations. They translated measured plunger overtravel and other machine deviations into corrected simulation boundary conditions, improving agreement with observed flow fronts and injected mass. The corrected simulations retained mass differences of 1.85–5.78 mg, averaging 3.60 mg. The numerical corrections are specific to this machine, POM material and dog-bone geometry; long-term overtravel variability was not investigated.

A machine is told to stop. Its piston keeps moving. For a plastic part just 12 mm long, that extra travel changes the story.

Explore how Trotta, Stampone and Tosello measured a real moulding machine—and brought its behaviour into a computer simulation.

The stop command isn’t the stopping point.

Choose a speed and a commanded stroke—the distance the piston should travel. Then reveal the extra motion measured in the experiment.

Short-shot calibration lab

Schematic replay · measured stroke data from Fig. 13. Filling shapes are a qualitative guide to Figs. 15–16; animation is slowed.

At 200 mm/s, the paper’s parts were already full at 8 mm. No 10 mm stroke breakdown was reported; showing 8 mm.

Measured contributions from Fig. 13. Total travel is their sum plus the commanded stroke.

A short shot interrupts injection before the final pressure-holding stage. The partly filled plastic shows where the flow has reached. Start by revealing the motion the machine adds after the commanded stroke.

A better model begins with the real machine.

Hot polyoxymethylene, or POM , is pushed through a runner into two tiny test pieces. A computer model predicts how that plastic fills the mould—but it needs the right inputs.

The study’s central move was to replace nominal settings with conditions informed by what the machine actually did: extra piston travel, decompression effects and temperature offsets.

Measure the motion

Monitor piston position, speed and pressure. The two-plunger machine’s inertia and drive response let its injection piston travel beyond the command.

Freeze the filling front

Stop shots at selected strokes. Partly filled specimens expose the path the plastic has taken, giving a visible comparison with the 3D simulation.

Calibrate, then compare

Use corrected simulation inputs and check both the shape of the front and the mass of the moulded pieces. The corrected fronts agreed more closely with the experiments.

Speed changes the clock. The commanded stroke sets the front.

In the paper’s nominal simulations, flow fronts at a fixed stroke stayed essentially similar across speeds and these temperature settings. Try the timing inputs below.

Table 2 values are reproduced as data; the front drawing and replay are independent schematics. These times belong to the nominal simulation and do not include the corrected travel.

Measure first. Simulate more faithfully.

The corrected model captured the observed filling fronts more closely. Its predicted masses were still higher than the measured masses, with an average remaining difference of 3.60 mg.

The practical implication: a machine’s real response belongs in its simulation. Monitoring and short shots provide a route from machine behaviour to better filling inputs.

Average remaining mass difference across the analysed short-shot conditions. The corrected simulations were heavier in every comparison.

A short shot deliberately leaves a mould partly filled. In this study, injection was interrupted before the pressure-holding stage used in normal production. The resulting plastic pieces reveal the filling front—the boundary between plastic and empty cavity. Comparing these pieces with simulations helps identify which model inputs need correction.

The moving piston and its drive cannot stop instantly. At higher injection speeds, the measured extra stroke became larger. Figure 13 separates a small switch-over contribution from a larger cushion-related contribution. At 8 mm commanded travel, their combined extra motion rises from 0.84 mm at 100 mm/s to 3.40 mm at 200 mm/s.

They monitored machine behaviour and produced POM specimens under varied speeds and temperatures. A targeted short-shot plan used 6, 8 and 10 mm nominal strokes. Separately, machine characterisation used eight combinations of melt temperature, mould temperature and speed. The paper compares measured motion, visible fronts and specimen masses with 3D numerical simulations.

The numerical corrections belong to the studied two-plunger machine, POM material and twin-specimen geometry. The workflow—monitor the machine, inspect short shots, correct the inputs, then compare fronts and mass—offers a way to obtain suitable corrections for another process when its own measurements are available.

Performance characterisation of a micro-injection moulding machine and influence of operative conditions on the process simulation accuracy

G. Trotta · B. Stampone · G. Tosello

Journal of Manufacturing Processes 172 (2026), 924–938. DOI: 10.1016/j.jmapro.2026.06.019

Visual and data sources. Geometry: Fig. 3 and Section 2.1. Piston travel: Fig. 13. Short-shot fronts: Figs. 15–16. Nominal filling times: Table 2. Corrected mass comparisons: Section 4.3 and Figs. 17–19.

The graphical abstract is an AI-generated conceptual illustration. Interactive drawings independently re-express the paper’s geometry and findings; they are schematic, with slowed animation. Travel totals are calculated from the published components.

Original article © 2026 The Authors, published by Elsevier on behalf of The Society of Manufacturing Engineers, under CC BY-NC-ND 4.0. The original paper figures remain available in the linked article.

*Calculated: commanded stroke + switch-over + cushion-related contribution. The source does not report a 10 mm / 200 mm/s stroke breakdown.